Basics of Shielding Design for Charged Particle Therapy

Basics of Shielding Design for Charged
Particle Therapy Facilities
p
n
π
K
Nisy Elizabeth Ipe, Ph.D., C.H.P.
Consultant, Shielding Design, Dosimetry
& Radiation Protection
San Carlos, CA, U.S.A.
Email: nisy@comcast.net
“Doing the right thing, doing it right”
PTCOG EW 2010, Nisy E. Ipe
1
PTCOG REPORT 1: Shielding Design and Radiation
Safety of Charged Particle Therapy Facilities
(http://ptcog.web.psi.ch/archive_reports.html)
1.
2.
3.
4.
5.
6.
7.
8.
Introduction – N. E. Ipe
Radiological Aspects of Charged Particle Therapy Facilities –
N. E. Ipe
Shielding Design Considerations – G. Fehrenbacher & N. E.
Ipe
Radiation Monitoring – Y. Uwamino & G. Fehrenbacher
Activation – Y. Uwamino
Monte Carlo Codes - S. Roesler
Patient Dose from Secondary Radiation – H. Paganetti & I.
Gudowska
Safety Systems and Interlocks – M. Schippers
Advisors: A. Smith, A. Mazal and D. Jones
EW 2010, Nisy E. Ipe
Consultants: S. Ban and PTCOG
H. Yashima
2
Secondary Radiation
• Produced by interaction of particles with beam
line components
• Produced at locations where beam losses occur
– In synchrotron and cyclotron during injection,
acceleration and extraction
– During energy degradation in cyclotron
– During beam transport to treatment room
– In beam shaping devices located in treatment nozzle
• Also produced in patient, dosimetric phantom
and beam stop
• Accelerators, beam transport line and treatment
rooms require shielding
PTCOG EW 2010, Nisy E. Ipe
3
Secondary Radiation from Protons
• Interaction of protons (50 MeV -1000 MeV)
with matter results in the production of an
intra-nuclear cascade (spray of charged and
neutral particles)
• Intra-nuclear cascade is important for protons in
therapeutic energy range of interest (67 MeV –
330 MeV)
• Five distinct stages
–
–
–
–
–
Intra-nuclear cascade
Production of muons
Electromagnetic cascade
Evaporation of nucleons and fragments
Activation
PTCOG EW 2010, Nisy E. Ipe
4
Intra-Nuclear Cascade
• Incoming hadron (p,n,..)
interacts with individual
nucleons in nucleus,
producing spray of particles
• Pions and kaons are
produced above ~ 135 MeV
and ~ 495 MeV, respectively
• Large fraction of energy
transferred to single
nucleon
p
n
• This nucleon with E>
150 MeV propagates
cascade
• Nucleons with energies
between 20 -150 MeV
transfer energy to
several nucleons (~ 10
MeV/nucleon)
• Charged particles are
quickly stopped by
ionization
• Neutrons predominate
π
K
PTCOG EW 2010, Nisy E. Ipe
5
Evaporation and Activation
•After interaction
with incoming hadron,
remnant of original
nucleus is left in an
excited state
n
• It de-excites emitting α
particles with typical
energy < 30 MeV
•Particles include
“evaporation
nucleons” (protons
and neutrons) and
some fragments
α
β
• Proton energy
deposited locally
n
• Evaporation neutrons
travel long distances
• Remaining excitation
p energy emitted as
gammas
γ • De-excited nucleus
may be radioactive
(activation)
PTCOG EW 2010, Nisy E. Ipe
6
Muons and Electromagnetic Cascade
•Charged pions decay to muons and
neutrinos
µ
π±
• Muons are penetrating particles,
and deposit energy by ionization,
•
photonuclear reactions also possible
±
e+
π0
γ
γ
en
• Neutral pions decay to gammas
which initiate electromagnetic (EM)
cascade
• Attenuation length of EM cascade
is much shorter than that for
neutrons
• Neutrons are principal
propagators of cascade with
PTCOG EW 2010, Nisy E. Ipe
increasing depth
7
Unshielded Neutron Spectra for 250 MeV
Protons Incident on Thick Fe Target (FLUKA)
Evaporation
Neutron Peak
PTCOG EW 2010, Nisy E. Ipe
Agosteo et al., NIM Phys. Res. B265, 581-589. 2007
8
Neutron Yields for Protons Incident on a
Thick Iron Target (FLUKA)
Proton Range Iron Target
Neutron Yield (n/p)
Energy, (mm)
Ep
Radius Thickness En
En >
Total
(MeV)
<19.6 19.6
(mm)
(mm)
MeV MeV
100
14.45
10
20
0.118
0.017
0.135
150
29.17
15
30
0.233
0.051
0.284
200
47.65
25
50
0.381
0.096
0.477
250
69.30
58
75
0.586
0.140
0.726
PTCOG EW 2010, Nisy E. Ipe
Agosteo et al., NIM Phys. Res. B265, 581-589. 2007
9
Characteristics of Shielded Neutron Field
• High-energy neutron component propagates cascade
in shield and continuously regenerates lower-energy
neutrons and charged particles at all depths in the
shield via inelastic reactions
• Yield of lower-energy neutrons increases as proton
energy increases
• Greater yield of lower-energy neutrons is more than
compensated for by greater attenuation in shield
because of higher cross-section at lower energy
• Radiation field reaches equilibrium condition beyond
a few mean free paths within shield
• Typical neutron spectrum observed outside a thick
shield consists of peaks at ~ few MeV and at ~ 100
MeV
PTCOG EW 2010, Nisy E. Ipe
10
Carbon Ion Interactions
• Nuclear interactions
occur by either grazing or
head-on collisions
• In grazing collision
fragmentation of incident
ion or target nucleus occurs
• Only segments of nuclei
that interpenetrate undergo
significant interaction and
mutual disintegration
•Evaporated nucleons and
light clusters are produced
on oxygen
12C
10C
16O
Grazing Collision
PTCOG EW 2010, Nisy E. Ipe
15O
11
Courtesy of Igor Pshenichnov
Carbon Ion Interactions
• Head-on collisions are less frequent but large energy
transfer occurs
• Projectile breaks into many small pieces and no
high-velocity fragment survives (unlike in grazing
collisions)
•Secondary particles are produced
on hydrogen
H
12C
9B
12C
16O
H
H
Li
PTCOG EW 2010, Nisy
E. Ipe
Head-on
Courtesy of Igor Pshenichnov
Collision
12
Relative Dose
Relative Dose at 1 m (0°-10°) from 430 MeV/u
Carbon Ions Incident on ICRU Tissue
Neutron
1.E+01
1.E+00
1.E-01
1.E-02
1.E-03
1.E-04
1.E-05
1.E-06
Proton
Photon
Charged
Pion
15
75
135 195 255 315 375 435 495
Concrete Thickness (cm)
PTCOG EW 2010, Nisy E. Ipe
N.E. Ipe and A. Fasso, SATIF8, 2006
13
Neutron Spectra in Concrete from 430 MeV/u
Carbon Ions Incident on ICRU Tissue
340 MeV
-2
Fluence (cm per carbon ion)
1.E-03
1.E-04
2.3 MeV
1.E-05
0-10 degrees: Concrete
Surface
80-100 degrees: Concrete
Depth = 4.8 m
0-10 degrees: Concrete
Depth = 4.8 m
80-100 degrees
1.E-06
1.E-07
145 MeV
145 MeV
1
4
5
1.E-08
1.E-09
1.E-10
1.E-11
1.E-12
1.E-13
1.E-14
1.E-04
1.E-03
1.E-02
1.E-01
1.E+00
1.E+01
Neutron Energy (GeV)
PTCOG EW 2010, Nisy E. Ipe
N.E. Ipe and A. Fasso, Proceedings of SATIF8, 2006, Korea
14
Secondary Radiation Field
• Quite complex
• For shielding, neutrons are the dominant
component
• Important to understand how neutrons
interact
PTCOG EW 2010, Nisy E. Ipe
15
Neutron Energy Classification
• Thermal:
Ēn = 0.025 eV at 20ºC
Typically
En ≤ 0.5 eV
• Intermediate : 0.5 eV <En ≤ 10 keV
• Fast:
10 keV< En ≤ 20 MeV
– Includes evaporation neutrons
• Relativistic
En > 20 MeV
PTCOG EW 2010, Nisy E. Ipe
16
Elastic Scatter
• Kinetic energy and
momentum are conserved
• Fast neutrons thermalized
by elastic scatter
• Interaction with hydrogen
is like a billiard ball
collision
• Primary process of energy
loss below 1 MeV in
hydrogenous materials
(concrete, polyethylene..)
• Dominant interaction below
10 MeV for all materials
Hydrogenous materials
are most efffective for
fast neutron shielding
n
Nucleus
http://www.glossary.oilfield.slb.com/Display.cfm?Term=elastic%20neutron%20scattering
PTCOG EW 2010, Nisy E. Ipe
17
Inelastic ScatterHigh-Z material used for shielding
• Kinetic energy is not
conserved
• Nucleus absorbs energy
and is left in an excited
state
• De-excites emitting
gamma rays
• Is dominant process
above 10 MeV in all
materials
• In high-Z materials,
inelastic scattering
reduces neutron energy,
thus making
hydrogenous material
that follows more
effective
must always be followed by a
hydrogenous material because highZ materials are transparent to lower
energy neutrons
Nucleus
Inelastic
gamma
rays
PTCOG EW 2010, Nisy E. Ipe
Nucleus
18
http://www.glossary.oilfield.slb.com/Display.cfm?Term=inelastic%20neutron%20scattering
Thermal Neutron Capture
• Thermal neutrons are
captured by the
nucleus.
Use borated polyethylene in
• The excited nucleus
emits capture gamma
maze door instead of
rays
polyethylene
• Capture cross section
(< 1 keV) decreases
with increasing
neutron energy
• 1H (n,γ)2H -2.2 MeV
(polyethylene &
Thermal
concrete)
Capture
Excited
neutron
• Boron has higher
Gamma
Nucleus
capture cross-section
Ray
and lower capture
gamma energy (0.478
MeV)
http://www.glossary.oilfield.slb.com/Display.cfm?Term=inelastic%20neutron%20scattering
PTCOG EW 2010, Nisy E. Ipe
19
Neutron Interactions (En< 20 MeV)
• Neutrons are uncharged and can travel
long distances
• Almost all interactions are scatters (elastic
or inelastic)
• Absorption important only at thermal
energies and in a few resonances in keV
region
PTCOG EW 2010, Nisy E. Ipe
20
Neutron Interactions (En> 20 MeV)
• Relativistic neutrons arise from cascade processes
(proton) and nuclear and fragmentation processes (ion)
• Important in propagating cascade
• Neutrons in cascade have energies as high as the primary
proton beam
• Neutrons from ion accelerators have maximum energy of
~ 2 x specific energy of ion (860 MeV for 430 MeV/u
carbon ion)
• High-energy neutron component (En > 100 MeV)
– propagates neutrons through shielding
– regenerates lower-energy neutrons and charged particles at all
depths via inelastic reaction
– charged particles are absorbed in the thick shielding
– lower-energy neutrons undergo capture reactions giving rise to
neutron-capture gamma rays
PTCOG EW 2010, Nisy E. Ipe
21
Neutron Interactions (En> 20 MeV)
• Low-energy component (50 MeV-100 MeV)
– Intra-nuclear cascade
– Evaporation nucleons
– Activation
PTCOG EW 2010, Nisy E. Ipe
22
Calculational Methods
1. Monte Carlo (MC) Codes
– FLUKA, MCNP, MCNPX, GEANT, etc.
– Full computer simulation modeling
accelerator, beam line and room geometry
– Can be used to generate isodose curves, thus
providing visualization of secondary
radiation field
– Can be used to derive computational models
– Time consuming
PTCOG EW 2010, Nisy E. Ipe
23
Calculational Methods
2. Analytical Methods
-
-
-
-
Most models are line-ofsight and assume point
source
Usually limited to
transverse shielding and
simple geometries
Don’t account for
changes in angle of
production, target
material and dimensions,
shielding material,
density and composition
Example: Moyer Model
derived for GeV proton
machines
 d
H
H=
exp − 
r
 λ
0
2
H = dose at point of interest
H0 = dose at 1 m from source
d = slant thickness
r = distance to shield
PTCOG EW 2010, Nisy E. Ipe
24
λ = attenuation
length
PTCOG
Report 1
Attenuation Length
• Attenuation length (λ) is the penetration distance in which
the intensity of radiation is reduced by a factor of e
• Measured in cm, or in g-cm-2
• λ changes with depth and reaches an equilibrium value
Nakamura, T. in Proc. Shielding Aspects of Accelerators, Targets and Irradiation Facilities (SATIF-7),
2004, NEA, OECD, Paris
Attenuation Length (g cm -2)
150
SLAC,NE213
HIMAC,TEPC
HIMAC,Cal.
ISIS,Rem C.
ISIS,Cal.
Loma Linda,TEPC
Orsay,TEPC
TIARA,NE213
CYRIC,NE213
100
50
0
1
10
2
10 E. Ipe
PTCOG EW 2010, Nisy
Maximum Source Neutron Energy (MeV)
3
10
25
Calculational Methods
3. Computational Models
– Comprised of source terms and attenuation
lengths as a function of angle
– Particle energy, angle of production, target
material, dimensions, shielding material,
composition and density are considered
– During schematic design phase, facility layout
undergoes several iterations
– Full Monte Carlo simulations for specific
room design is time consuming and not very
cost effective
PTCOG EW 2010, Nisy E. Ipe
26
– Computational
models are useful at this stage
H 0 ( E p ,θ )
r
2

d 
exp −

 λθ g (θ ) 
Computational Models

d 
H( Ep, θ, d/λg(θ))H=0 ( E p ,θ )
exp −

2
r
 λθ g (θ ) 
Where:
H is the dose equivalent at the outside the shield,
H0 is source term at an angle θ with respect to the incident beam, and
is assumed to be geometry independent
r is the distance between the target and the point at which the dose
equivalent is scored,
d is the thickness of the shield
d/g(θ) is the slant thickness of the shield at an angle θ
λθ is the attenuation length at an angle θ
g(θ) is cosθ for forward shielding
g(θ) is sinθ for lateral shielding
g(θ) = 1 for spherical geometry
PTCOG EW 2010, Nisy E. Ipe
27
Dose From 430 MeV/u Carbon Ion Incident
On 30 cm ICRU Tissue Sphere
Dose Per Carbon ion @ 1 m: 0 to 10 degrees
Vacuum
pSv-m2/ion
1.E+02
Concrete
1.E+00
30 cm Fe +Concrete
1.E-02
60 cm Fe + Concrete
1.E-04
90 cm Fe + Concrete
1.E-06
15
75
135 195 255 315 375 435 495
120 cm Fe + Concrete
Shielding Thickness (cm)
N.E. Ipe and A. Fasso, SATIF8, 2006
PTCOG EW 2010, Nisy E. Ipe
28
Computational Models for Composite Shield (Iron +
Concrete) Thickness > 1.35 m
430 MeV/u Carbon on Tissue
Iron
(cm)
0-10°
0
H0
(pSvm2/ion)
3.02±.04
30
1.25±0.02
60
6.05±0.03
x x1.E-01
90
2.77±0.09
x1.E-01
1.33±0.05
x 1.E-01
120
10-30°
λ (g/cm2)
H0
λ (g/cm2)
(pSvm2/ion)
123.81±0.48 4.81±0.06 133.09±0.7
4
x1.E-01
123.12±0.38 2.44±0.03 129.64±0.3
6
x1.E-01
120.32±0.46 1.11±0.04 128.66±0.7
0
E-01
119.58±1.25 5.27±0.29 126.09±0.8
0
x1.E-02
117.68±0.91 2.48±0.24 124.29±0.9
4
x1.E-02
PTCOG
EW 2010, Nisy E. Ipe
Ipe & Fasso, SATIF8, 2006
40-60°
H0
(pSvm2/ion)
4.71±0.21
x1.E-01
1.91±0.08
x1.E-01
8.29±0.66
x1.E-02
3.29±0.69
x1.E-02
1.34±0.68
x1.E-02
λ (g/cm2)
117.64±1.32
119.38±0.48
118.5±0.80
119.14±1.34
118.83±2.89
29
Total Dose Equivalent Transmission Curves For 430 MeV/u
Carbon Ion Incident 0n Tissue
0 to 10 degrees
Concrete
30 cm Iron + Concrete
Transmission
1.0E+01
1.0E+00
60 cm Iron + Conrete
90 cm Iron + Concrete
1.0E-01
1.0E-02
120 cm Iron + Concrete
4.65 m
concrete
1.0E-03
1.0E-04
1.0E-05
0
3.3 m
+ 165 cm concrete Proshield 300
100
200
ProShield Ledite 300
Ledite 293
120 cm iron
1.0E-06
1.0E-07
Iron
300
400
Ledite 247
500
Shielding Thickness (cm)
N.E. Ipe, ICRS11, 2008
PTCOG EW 2010, Nisy E. Ipe
30
Shielding Design Considerations
• Accelerator Type
– Synchrotron
– Cyclotron
– Self shielding effects of beam line components
• Type of Shielding Material
– Composition
– Density
– Water content
• Facility Layout
– Adjacent occupancies
– Type of Area (Controlled, Public, etc.)
– Above ground, underground..
• Country/State Specific Regulatory Dose Limit
Shielding design is facility dependent!
PTCOG EW 2010, Nisy E. Ipe
31
Shielding Design Considerations
• Treatment
– Particle type (proton, carbon ion, …)
– Energies
– Current or intensity at each energy
– Beam shaping and delivery (scanning vs. scattering,
etc.)
– No. of patients/year
– No. of fractions/patient at each energy
– Dose delivered per fraction
– Beam-on time
– Beam losses and locations
– Target materials and dimensions
Treatment parameters vary from facility to facility
PTCOG EW 2010, Nisy E. Ipe
32
Fixed Beam Room Shielding Considerations
•
•
am
Be
•
l
ica
rt
Ve
•
°
•
45
•
Horizontal and 45 °
vertical beams
Use Factor = 1/2 for each
beam, or 2/3 and 1/3
Neutrons from carbon ions
dominate shielding in
forward direction
Neutrons from protons
dominate at large angles and
at maze entrance when
proton intensity is higher
No shielded door at maze
entrance
Consider contributions from
adjacent areas
Horizontal
Be am
Courtesy of Siemens Medical Solutions U.S.A, Inc.
PTCOG EW 2010, Nisy E. Ipe
33
Gantry Room Shielding Considerations
• Beam rotates
around patient
• Beam stopper is
asymmetric and
rotates with beam
• Ceiling, lateral
walls and floor see
forward radiation
• Walls in forward
direction thinner
than Fixed Beam
Room because of
lower Use Factor
(1/4)
Courtesy of Siemens Medical Solutions U.S.A, Inc.
PTCOG EW 2010, Nisy E. Ipe
34
Skyshine and Groundshine (PTCOG Report 1)
PTCOG EW 2010, Nisy E. Ipe
35
Ducts/Penetrations (PTGOG Report 1)
Extension
of Duct
Length
Curved
Bend
One Bend
Two Bends
PTCOG EW 2010, Nisy E. Ipe
Shadow
Mask
Shield
36
Shielding Thicknesses For Particle Therapy Facilities
(PTCOG Report 1)
CNAO,
Italy
Gunma,
PTCOG
EW 2010, Nisy E. Ipe
Japan
NCC,
Korea
PSI,
Switzerland
37
Architect’s 3-D Rendition of an SRS Monarch
250 Single-Room Facility
PTCOG EWAn
2010,
Nisy E.
Ipe
38
Courtesy of The Benham Companies,
SAIC
Company,
Oklahoma City, Oklahoma
CONCLUSIONS
“One Size Fits All” Approach to Shielding
DOES NOT WORK
NCCHPS BOARD OF DIRECTORS 2005
Shielding is facility specific
PTCOG EW 2010, Nisy E. Ipe
39
PTCOG EW 2010, Nisy E. Ipe
40